►
From YouTube: 🖧 IPLD Every-two-weeks Sync 🙌🏽 2022-11-21
Description
An every two weeks meeting to sync up on all IPLD (https://ipld.io) related topics. It's open for everyone and recorded. https://github.com/ipld/team-mgmt
A
B
Welcome
everyone
to
this
week's
ipld
meeting,
it's
November,
the
21st
and
as
every
two
weeks
we
go
over
the
stuff
that
people
have
worked
on
and
then
discuss
any
open,
Agenda
items
or
answer
any
questions.
Anyone
might
have
so.
B
For
the
weekly
updates,
I
don't
have
an
update.
I
certainly
didn't
do
any
iPod
work
last
week,
although
there
will
be
a
raster
apld
release
soon,
which
I
really
should
do
in
the
notes.
I,
don't
see
any
updates
from
anyone
else.
C
I
I,
don't
actually
have
anything
worth
talking
about
either
the
only
the
only
thing
I
I
wanted
to
flag
was
I'm
going
to
I.
Haven't
figured
exactly
how
to
do
it
yet,
but
I'm
going
to
propose.
C
Add
change
to
cids
with
the
I've
talked
about
this
publicly
about
identity,
zids,
to
put
a
hard
limit
on
the
creation
and
decoding
of
identity,
CID
links,
so
I
I'm,
not
sure
whether
this
is
a
spec
thing
or
whether
it's
an
implementation
thing
or
both
so
I
might
I
might
do
it
in
the
spec
first,
if
there's
a
spot
for
it,
and
the
number
that
we've
sort
of
circled
around
was
2K
I'm
tempted
to
pull
it
down
to
1K.
But
we'll
do
that
in
a
discussion
on
GitHub.
C
But
that's
that's
pretty
much.
The
only
thing
that
is
sort
of
floating
around
for
me
at
the
moment.
D
Thanks
Rod
I
was
quick,
quick
question.
What
was
the
what
prompted
the
putting
a
limit
on
the
identity.
C
Because
we
we
keep
on
having,
we
keep
on
encountering
problems
bugs
even
security
issues
in
our
stack
caused
by
identity,
cids
and
the
fact
that
they're
not
limited
the
the
ability
to
create
arbitrarily
long
bite
strings
of
these
things
and
have
them
valid
and
then
be
able
to
use
them
for
various
things
creates
all
just
all
sorts
of
problems,
and
everyone
keeps
on
encountering
them
all
over
in
surprising
ways.
C
B
C
The
beginning,
so
the
idea
here
is,
is
let's
impose
a
limit.
Try
and
you
know,
we've
had
a
lot
of
discussion
about
what
a
limit
could
be
and
then
we
can.
You
know
if
they,
if
that
limit,
actually
ends
up
causing
a
problem
for
people
that
haven't
been
that
haven't
weren't
aware
of
the
discussions,
then
we
can
sort
of
negotiate
that
negotiate
that
at
the
exact
limit,
depending
on
situations,
yeah.
C
B
All
right
does
anyone
else
have
any
updates
or
questions
or
things
to
discuss.
C
E
For
the
review
rod
on
the
debugging
form
of
the
car,
PR-
that's
merged,
so
we
do
have
that.
E
That
did
you
I
think
we
did
last
time
yeah,
but
but
we
got
the
the
rest
of
that
work.
This
is
so
that
you
can
view
a
car
file
in
its
Json
form
sort
of,
but
in
a
human
readable
file,
and
then
you
can
recompile
it
back
into
the
binary
car
file
format
that
it
was
originally
in
to
allow
for
sort
of
manual
introspection,
and
then
you
can
sort
of
munge
change
that
that
Json
and
then
recompile
a
slightly
modified
version.
B
I
I've
spent
like
the
week
before
I've
spent
a
bit
on
time
on
drafting
the
parameterized
multi-hash
stuff
and
I
wanted
to
use
Poseidon
as
an
example
and
then
figure
out
that
Poseidon
has
more
parameters
than
I
thought.
It
has
so
because,
like
for
ficon,
we
use
Poseidon
and
we
specify
an
RT
and
we
specify
one
more
parameter.
B
But
then
I
found
out
oh
there's
more
parameters
that
we
actually
have
hard-coded
for
Phi
coin,
but
are
generally,
if
you
read
the
Poseidon
paper
and
you
have
more
parameters,
and
you
kind
of
like
have
a
python
script
to
calculate
which
parameters
you
want
to
use
for
your
security
guarantees
and
so
on.
So
anyway,
I
think
it's
like
it's
good
to
dig
into
so.
For
me,
it
was
interesting
and
also
like
it's
good
to
know
those
things
and
yeah.
It
even
makes
it
more
worth.
B
Having
like
some
way
for
pyramidal
tries
multi-hash,
because
it's
just
like
yeah,
it's
basically
Unbound
more
or
less
like
you
can
really
have
like
an
infinite
amount
of
possible
combinations
of
those
parameters
and
yeah
anyway.
So
I've
yeah
I
yeah
I,
want
to
spend
some
time
on
this
because
it's
kind
of
related
to
what
I
also
have
in
draft
mode
is
the
whole
like
CAD
version
2,
which
we
discussed
at
that
week.
B
But
it's
like
the
why
all
this
came
up
is
like
it's
so
tightly
connected
that
I,
yeah
I
need
to
start
somewhere
and
I
had
hope
that
the
pyramidalized
multi
hair
stuff
would
be
like
a
quick
one,
but
obviously
it's
not
yeah
but
yeah,
but
hopefully
we
will
get
to
the
CID
version,
2
stuff
as
well
yeah.
So
I
still
believe
that
all
cases
that
we've
talked
about
are
still
handled
by
all
this
and
I
will
link
in
the
notes
to
the
CID
version
to
stuff
to
the
current
one.
B
In
case
someone
is
interested
in
the
stuff
because
it's
floating
around
India
still
worth
having
a
look.
Yeah
am
I
here
like
in
the
chat.
Dimitri
says
that
you're
excited
about
the
perimeterized
matri
hash,
so
yeah
I
wanted
to
ask
like
what's
yeah?
Okay,
you
have
a
link.
So
what's
your
background?
Can
you
quickly
so.
D
B
D
So
special
hi
everybody
I'm.
Definitely
a
fan
of
ipld
I'm,
trying
to
figure
out
more
about
it.
I
was
actually
a
co-worker
of
Nikola
greca,
who
I
think
was:
oh
one
of
the
people
working
on
ipld
earlier,
but
I'm
I'm,
coming
from
the
decentralized
identifiers,
verifiable
credentials
and
encrypted
storage
lab
and
there's
a
number
of
things
that
we
use.
D
So
that
we
can
hash
link
so
that
we
can
use
the
same
nice
CID
mechanism
they
use
in
iPod,
but
in
arbitrary
Json
in
with
verify
the
credentials
so
that
we
can
cryptographically
bind
multiple
Standalone
find
credentials
to
each
other.
And
so
in
the
process
of
doing
that.
So
we
needed
essentially
Digest
multi-base,
which.
D
The
multi-hash
mechanism
would
be
almost
perfect
for,
but
we
have
one
additional
parameter.
We
have
an
optional
transform
parameter
so
before
applying
the
hash
on
an
object,
you
transform
the
object.
So
much
like
you
do
in
I
forget
what
it
was
called
dad
seabar,
where
you
have
to
sort
the
Json
queue
basically,
and
so
we
needed
that
additional
parameter
in
that
PR
261.
D
We
ended
up
adding
values
to
the
iPod
tag
of
the
multi-codac
list,
not
to
the
multihash.
Just
because
again
we
need
room
for
an
extra
parameter.
That
was
the
transform.
B
B
What
what
so,
basically
I've
discussed
it
with
the
people
from
the
lurk
project
where
they
have
like
they
do
like
cryptography
stuff,
and
it's
like
it's
hashing
on
an
elliptic
curve
and
it's
kind
of
like
it
is
hashing,
but
it's
kind
of
like
it's
kind
of
like
so
it's
kind
of
addressed
but
kind
of
like
a
different
world
from
what
we're
used
to
I
would
say,
and
they
also
kind
of
like
need,
but
so
I
think
it's
more
like
a
traditional
case.
B
We
basically
have
just
really
like
a
family
of
hash
functions
in
this
case
Poseidon,
and
you
have
parameters,
but
I
think
your
case
could
also
be
applied.
So
just
what
I'm.
Currently
my
thinking
is
or
like,
where
I
come
to
with
this
discussions
with
others.
Is
that
you
kind
of
specify
hash
family.
So,
in
my
case
it
would
be
Poseidon
and
then
you
specify
have
just
a
structured
object,
which
probably
is
something
like
Json,
probably
and
there
you
can
specify
parameters,
and
you
say
it
for
this
family.
B
The
idea
is
that
you
don't
use
huge
hashes,
but
you
really
use
smaller
ones,
something
in
the
range
of
like
32
was
64
bits
only
and
because
having
clashes
or
there
from
your
hedge
is
not
a
problem,
because
you
really
you
want
to
use
them
as
an
identifier.
But
you
want
to
have
like
a
nice
way
to
get
to
an
identifier.
So,
instead
of
memorating
them
you
just
hash
them
and
and
in
case
you
would
have
a
collision
with
your
hash.
B
You
can
just
add
a
salt
to
it
and
and
end
up
with
a
different
hash.
But
it's
kind
of
like
the
kind
of
prefix.
Is
your
hash
family
like
Poseidon,
and
then
you
have
those
hashes
of
your
perimeters
and
again
it
could
end
up
in
a
table
or
not,
but
probably
it's
a
table
for
the
well-known
ones
because,
like
although
the
space
is
huge,
probably
people
practically
will
only
use
a
small
subset
of
those.
D
Yep
that
makes
sense
and
I
guess.
My
only
question
is
that
Json
object
that
you're
hashing.
Do
you
plan
to
sort
it
or
do
you
want
to
require
sorting
right
just
so
that.
B
Yes,
so
my
idea
was
to
get
it
I'm
actually
regretting
this
down
and
it
will
be
like
totally
open
for
discussion
like
we
can
change
everything,
but
my
idea
would
be
it.
I
specified
it
as
deck
Chase
on
without
links.
So
it's
basically
just
it's
I
was
specified
as
Jason
without
white
space
and
keys
sorted
by
by
by
bytes.
Basically,
so
you
sort
the
keys,
you
don't
have
any
right
space
and
it's
Json.
B
Basically,
this
would
be
my
idea
and
and
as
you
basically
with
your
additional
multi-coding
like
hash
family
What,
I
Call
Family,
you
kind
of
like
specify
there,
you
specify
which
keys
are
available
so
that
you
don't
end
up
with
like
slightly
different
key
names
like
some
people.
They
call
it
like
impossible
called
erity
or
you
call
it
only
a
or
something,
but
you
should
basically
have
a
well
specified
format,
but
it's
like
it's
really
like
lightweight,
so
I
wouldn't
expect.
B
B
That's
the
basic
idea:
yeah
cool
yeah
and
yeah,
so
I
yeah
I
should
really
work
on
this
one
cool,
but
it's
cool
to
hear
that
it
seems
to
also
match
other
people's
use
cases,
because
this
is
the
whole
point
like
we
finally
get
the
this
nail
down
because,
like
different
people
ask
for
it
so
cool
all
right
anything
else.
C
Probably
just
worth
reminding
again
so
move
is
trying
to
organize
a
working
group,
iplally
working
group,
just
interested
parties
providing
input
on
prioritization
for
ipld
work
and
connecting
with
other
people,
doing
ipld
things
trying
to
find
people
doing
common
things
and
connecting
them
together.
So
there's
an
issue
for
that
which
I
don't
have
the
top
of
my
head,
so
Moe's,
currently
trying
to
figure
out
timing
and
and
people
so
just
wrapped
the
link
to
the
right.
It's
okay,
yeah!
It's
issue,
number
249
in
the
ipld
ipld
repo!
C
A
During
the
I
think,
I
had
a
covet
or
something
but
yeah
just
wanted
to
chime
in
say,
hi.
That's
all
really.
B
Cool
yeah,
it's
yeah.
For
me,
it's
cool
to
see
more
people
are
showing
up.
Obviously
so
that's
cool
all
right,
but
if
there
isn't
anything
else,
I
think
we
just
closed
the
meeting
for
this
week
and
yeah
see
you
all
again
in
two
weeks,
then,
and
of
course,
so
we
have
an
after
party
so
yeah
you
can
keep
staying
and
yeah.
So
hey
goodbye.
Everyone
bye.